CN101184013B - Method for preventing generation of loop, host node and system - Google Patents

Abstract

The invention discloses a method for avoiding loops. The method includes: a master node on an Ethernet ring network sends a health detection message on a master port and a slave port, and if only one of the master port and the slave port receives a health detection message, the master node makes the secondary port in a blocked state. The invention also discloses a master node and a system for avoiding loops. The invention avoids the shortcoming of a loop when a single-pass failure occurs in the Ethernet ring network.

Description

A method, master node and system for avoiding loops

technical field

The invention relates to the Ethernet ring network technology, in particular to a method, a master node and a system for avoiding loops. the

Background technique

FIG. 1 is a schematic structural diagram of a single Ethernet ring network. Referring to Fig. 1, a single Ethernet ring network includes a master node and multiple transmission nodes, and the master node and each transmission node transmit service data on the Ethernet ring network to complete various service functions. the

Referring to FIG. 1 , since the Ethernet ring network is a ring network structure, it is necessary to prevent loops, thereby avoiding broadcast storms caused by unknown unicasts and broadcasts being circulated on the ring. In order to avoid loops in the Ethernet ring network, under normal conditions, the master node only opens the primary port on one side of itself, and blocks the secondary port on the other side of itself. In this way, when transmitting service data, the Ethernet ring network is a chain without loops. the

The master node needs to be responsible for monitoring the status of each transmission node and link on the Ethernet ring network to ensure the normal operation of the Ethernet ring network. At present, the process of ensuring the normal operation of the Ethernet ring network includes: the master node periodically sends health detection messages through the main port, usually called HELLO messages. If the master node can receive When the HELLO message is received, the master node determines that all the transmission nodes and links on the current Ethernet ring network are normal, so there is no need to troubleshoot, that is, the master node keeps the main port open and the secondary port is blocked; when the ring network fails , the transit node that senses the fault will report a fault notification message to the master node. After receiving the fault notification message, the master node determines that a transit node or link on the current Ethernet ring network is unavailable. Therefore, fault handling is required. That is, the primary node opens the secondary port. the

In actual service implementation, some ring network failures are not caused by link status changes. At this time, the port status of each node does not change, so the failure cannot be directly perceived. At this time, after the master node periodically sends the HELLO message from the primary port, it cannot receive the HELLO message from the secondary port within the specified three consecutive periods to sense the ring network fault. The cause of this situation may be the single-pass failure of the optical fiber link between two nodes in the Ethernet ring network, or the failure of the switching chip inside a certain node. In both cases there is a possibility of a single-pass failure. A unidirectional failure refers to a bidirectional link between two nodes in which the link in one direction is normal and the link in the other direction is faulty. For example, referring to Figure 2, in the bidirectional link between transit node 2 and transit node 3, forward link 1 fails, while reverse link 2 is normal (for the convenience of description, in this article, forward refers to The direction from the main port of the master node to the secondary port along the ring network, the reverse direction refers to the direction from the secondary port of the master node to the main port along the ring network), at this time, there is a forward disconnection of the Ethernet ring network and a reverse For single-pass failure of the path, when the master node sends a HELLO packet from the primary port, it cannot receive it from the secondary port.

Because in the prior art, as long as the master node does not receive the health detection message sent by itself, that is, the HELLO message, the primary port and the secondary port will be opened at the same time. Therefore, when a single-way failure occurs, the Ethernet ring network A loop in one direction is generated, and a broadcast storm caused by unknown unicast and broadcast transmission on the loop occurs. the

Contents of the invention

The object of the present invention is to provide a method for avoiding loops, a master node, and a system for avoiding loops, so as to avoid loops in the Ethernet ring network. the

In order to achieve the above object, technical scheme of the present invention is achieved in that:

A method for avoiding loops, applied to an Ethernet ring network with a ring network structure, the Ethernet ring includes a master node and more than one transmission node, and the master node and each transmission node transmit business data on the Ethernet ring network , to complete various business functions, and the master node is responsible for monitoring the status of each transmission node and link on the Ethernet ring network, the method includes:

The master node on the Ethernet ring network sends health detection packets on the primary port and the secondary port. , the master node keeps the secondary port in a blocked state;

Wherein, the step of the primary node sending the health detection message on the primary port and the secondary port includes: the primary node is in a complete state, the secondary port is blocked, and the primary node periodically sends the health detection message on the primary port. If the health detection message from the main port is not received on the secondary port, the health detection message is sent on the secondary port;

The health detection message received on only one of the main port and the secondary port is: the health detection message from the main port is not received on the secondary port, and the health detection message from the secondary port is received on the primary port ;

The master node making the secondary port in the blocked state includes: the master node keeps the secondary port in the blocked state. the

A master node is located in an Ethernet ring network with a ring network structure. The Ethernet ring includes a master node and more than one transmission node. The master node and each transmission node transmit business data on the Ethernet ring network to complete various Business functions, and the master node is responsible for monitoring the status of each transmission node and link on the Ethernet ring network, including:

The loop detection unit is used to notify the secondary port control unit to block the secondary port when switching to the complete state, and periodically send health detection packets on the primary port. If the secondary port does not receive the health detection packet from the primary port If the health detection message is sent on the secondary port, if the health detection message from the secondary port is received on the primary port, it is determined that the health detection message is only received on one of the primary port and the secondary port. Detect the message, and perform the processing of sending the blocking notification to the secondary port control unit;

The secondary port control unit is configured to make the secondary port in a blocking state after receiving the blocking notification. the

A system for avoiding loops, the system includes multiple transmission nodes and the master node proposed by the present invention. the

It can be seen that, in the present invention, the master node sends the health detection message on the two ports of the primary port and the secondary port, so if only one port receives the health detection message from another port, then it can be It is determined that a single-pass fault occurs on the Ethernet ring network, and the primary port blocks the secondary port. In this way, the disadvantage of a loop in the prior art when a single-way fault occurs in the Ethernet ring network is avoided, thereby avoiding a broadcast storm caused by unknown unicasts and broadcasts being transmitted cyclically on the loop. the

In addition, in the present invention, since the health detection message of the existing protocol supported by the device is used to detect whether there is a single-pass failure in the Ethernet ring network, there is no need to perform a complicated upgrade process on the device, which facilitates the realization and promotion of services . the

Description of drawings

FIG. 1 is a schematic structural diagram of a single Ethernet ring network. the

FIG. 2 is a schematic diagram of a single-pass fault occurring in an Ethernet ring network. the

Fig. 3 is a flow chart of avoiding loops in one embodiment of the present invention. the

Fig. 4 is a schematic structural diagram of a master node on an Ethernet ring network in an embodiment of the present invention. the

Detailed ways

In order to make the purpose, technical solution and advantages of the present invention clearer, the present invention will be further described in detail below in conjunction with the accompanying drawings and specific embodiments. the

The present invention proposes a method for avoiding loops. In this method, the master node on the Ethernet ring network sends a health detection message on the primary port and the secondary port, and if only one of the primary port and the secondary port receives the health detection message, the primary node blocks the secondary port. the

Fig. 3 is a flow chart of avoiding loops in one embodiment of the present invention. Referring to Fig. 3, in one embodiment of the present invention, the process of avoiding a loop specifically includes the following steps:

Step 301: When the master node is in a complete state (COMPLETE STATE), the master node blocks the secondary port and periodically sends a health detection message from the master port. the

In the prior art, the complete state of the master node has been defined, and the processing of this step is the same as the corresponding process in the prior art. the

It should be noted that, as in the prior art, in this process, when the Ethernet ring network initially works normally, that is, when the master node is in a complete state, each transmission node will monitor the status of the Ethernet ring network link. If a transmission node After monitoring an adjacent link failure, that is, a breakpoint in the Ethernet ring network, it will send a failure notification to the master node. After receiving the fault notification, the master node in the complete state can determine that the Ethernet ring network has a disconnection, and it will switch To the link failure state already defined in the prior art, when switching to this state, the master node opens the secondary port, and still periodically sends health detection messages from the master port. the

Step 302: When in the complete state, the master node judges whether the health detection message from the master port can be received from the slave port, if yes, return to step 301, otherwise, execute step 303. the

The way for the master node to judge whether it can receive the health detection message from the main port from the secondary port is to judge whether it can receive the health detection message from the main port from the secondary port in multiple consecutive cycles such as 3 cycles. the

In this step, the master node can determine the state of the forward link of the Ethernet ring network by judging whether the health detection message from the master port can be received from the slave port. When it cannot be received, it can be determined that the Ethernet ring network is disconnected in the forward direction. According to the processing method of the prior art, because the master node is in a complete state, its secondary port is blocked. Therefore, in order to ensure the transmission of subsequent business data, the master node should open the secondary port. port. However, in this case, since only the disconnection of the forward link of the Ethernet ring network has been determined, and the status of the reverse link of the Ethernet ring network has not been determined, it is necessary to perform subsequent steps (immediately send a health check in the reverse direction) message), so as to decide whether to open the secondary port according to the determined state of the reverse link of the Ethernet ring network. the

When it can be received, the forward path of the Ethernet ring network can be determined. Since the current master node is in a complete state, its secondary port is blocked, and the Ethernet ring network will not generate a loop. Therefore, you can return to step 301 and continue to detect the Ethernet ring network. Forward link state process. the

Step 303: The master node sends a health detection message from the secondary port, and judges whether the health detection message from the secondary port can be received from the primary port. If yes, perform step 304; otherwise, perform step 307. the

The manner of judging whether the health detection message from the secondary port can be received from the primary port may be: judging whether the health detection message from the secondary port is received from the primary port within one cycle. Here, only judging whether a health detection message from the secondary port is received within one cycle is to minimize the number of detections and improve the processing speed, so that when a single-pass failure occurs, it can respond quickly. the

In this step, the master node can determine the state of the reverse link of the Ethernet ring network by judging whether it can receive the health detection message from the secondary port from the master port. Since the forward link of the Ethernet ring network has been determined to be disconnected when this step is performed, if the health detection message from the secondary port cannot be received on the primary port, that is, the reverse circuit of the Ethernet ring network is disconnected, then it can be determined There is a breakpoint in the Ethernet ring network. In order to ensure the transmission of subsequent business data, it is necessary to perform the processing of the master node opening the secondary port in the subsequent step 307; if the health detection message from the secondary port can be received on the primary port, that is, the Ethernet ring network Reverse path, then, it can be determined that a single-pass failure occurs in the Ethernet ring network, and the processing of opening the secondary port in the prior art cannot be performed, otherwise, a loop will be caused on the reverse link of the Ethernet ring network, so continue to execute step 304 . the

Step 304: The master node switches to a pre-defined link single-pass state (SINGLE-PASSSTATE). the

In the prior art, only the above complete state and link failure state are defined for the master node. Moreover, through the processing of the above steps 301 to 303, it is possible to detect whether a single-way failure occurs in the Ethernet ring network when the master node is in a complete state. Preferably, in order to ensure that the master node can also detect whether a single-pass failure occurs in the Ethernet ring network under other circumstances, in this embodiment, the master node defines the link single-pass state in this step and the following step 307 The polling error status of . the

Step 305: When the link is in the single-pass state, the master node blocks the secondary port, and periodically sends health detection messages on the primary port and the secondary port respectively. the

Step 306: When the link is in the single-pass state, the master node judges whether the health detection message can be received from the primary port and the secondary port respectively. If the health detection message can be received on both the primary port and the secondary port, then Switch to the complete state, return to step 301, if no health detection message is received on the primary port and the secondary port, then perform step 307, if only one of the primary port and the secondary port receives the health detection message , return to step 305. the

In the above step 305 and step 306, when the master node is in the link unidirectional state, the master node needs to block the secondary port to avoid loops in the Ethernet ring network. Preferably, when the master node is in the link single-pass state, the master node also periodically sends health detection messages on the primary port and the secondary port respectively, in order to generate When changes, the status of the primary node and whether the secondary port is blocked can be adjusted in time. the

If the health detection message can be received on both the primary port and the secondary port, it means that the forward link and the reverse link of the Ethernet ring network are all paths, and the Ethernet ring network has returned to normal operation. Therefore, it is necessary to switch to complete state. If no health detection message is received on the primary port and the secondary port, it means that there is a breakpoint in the Ethernet ring network. In order to ensure the transmission of subsequent business data, the master node should perform the processing of the master node opening the secondary port in the subsequent steps. . If only one of the primary port and the secondary port receives the health detection message, it means that there is still a unidirectional fault in the Ethernet ring network, that is to say, the current state of the master node is still the link unidirectional state, so it should return Step 305 continues to execute the processing when the link is in the single-pass state. the

Step 307: The master node switches to a predefined polling error state. the

Step 308: When in the polling error state, the master node opens the secondary port, and periodically sends health detection messages on the primary port and the secondary port respectively. the

Step 309: When in the polling error state, the master node judges whether the health detection message can be received from the primary port and the secondary port respectively, and if the health detection message can be received on both the primary port and the secondary port, switch To the complete state, return to step 301, if a health detection message is only received on one of the primary port and the secondary port, then return to step 304, if no health detection message is received on the primary port and the secondary port , return to step 308. the

In the above step 307 to step 309, when the master node is in the polling error state, both directions of the Ethernet ring network are disconnected, and the master node needs to open the secondary port to ensure the transmission of business data. Preferably, when the master node is in the polling error state, the master node also periodically sends health detection messages on the primary port and the secondary port respectively, in order to change the forward link or reverse link of the Ethernet ring network. After that, the state of the primary node and whether the secondary port is blocked can be adjusted in time. the

If the health detection message can be received on both the primary port and the secondary port, it means that the forward link and the reverse link of the Ethernet ring network are all paths, and the Ethernet ring network has returned to normal operation. Therefore, it is necessary to switch to complete state. If the health detection message is received on only one of the primary port and the secondary port, then it indicates that a single-pass fault occurs in the Ethernet ring network, and it is necessary to turn to the process of avoiding loops in the Ethernet ring network in step 304 . If the health detection message is not received on the primary port and the secondary port, it means that there is still a polling error in the Ethernet ring network, that is to say, the current state of the master node is still in the polling error state, so it should return to step 308 Processing continues when polling for error status. the

It should be noted that the process shown in FIG. 3 above is only a preferred embodiment of the present invention. Those skilled in the art can understand that, in the case that the master node can work normally, the master node does not need to be limited to the consideration of the master node state and the Ethernet ring network link state in the above-mentioned Figure 3, and can execute on the master port and the slave port at any time. The process of sending health detection messages and judging whether the health detection messages are received on only one port to determine the single-pass failure of the Ethernet ring network, so that the secondary port is in a blocked state, and the processing of avoiding loops in the Ethernet ring network. For example, after the master node detects that the health detection message from the main port is not received on the secondary port and opens the secondary port, the master node periodically sends the health detection message on the primary port and the secondary port respectively; When a health detection message is received on a port, the master node performs the processing of blocking the secondary port. the

It can be understood that, although in the above embodiments, the steps of the method are described sequentially for ease of understanding, it should be noted that there is no strict limitation on the sequence of the above steps. the

Those of ordinary skill in the art can also understand that all or part of the steps in the above embodiments can be implemented by instructing related hardware through a program, and the program can be stored in a computer-readable storage medium, and the storage medium , can be ROM/RAM, disk, CD, etc. the

Correspondingly, the present invention also proposes a master node. Fig. 4 is a schematic structural diagram of a master node on an Ethernet ring network in an embodiment of the present invention. See Figure 4, the master node includes:

The loop detection unit is used to send a health detection message through the primary port and the secondary port, and if it is detected that the health detection message is only received on one of the primary port and the secondary port, a blocking notification is sent to the secondary port control unit;

The secondary port control unit is configured to block the secondary port after receiving the blocking notification. the

Preferably, the above-mentioned processing of the loop detection unit may specifically be: when switching to the complete state, notify the secondary port control unit to block the secondary port, and periodically send health detection messages on the primary port, if the secondary port If the health detection message from the primary port is not received on the main port, the health detection message is sent on the secondary port. Handling of notifications. the

Preferably, the loop detection unit is further configured to receive a health detection message from the secondary port on the primary port if the secondary port does not receive a health detection message from the primary port when it is in a complete state , then switch to the link single-pass state, after switching to the link single-pass state, perform the processing of sending the blocking notification to the secondary port control unit, and further when in the link single-pass state, between the main port and the The secondary port periodically sends the health detection message, and if the primary port and the secondary port both receive the health detection message, then switch to the complete state. the

Preferably, the loop detection unit is further configured to switch to the polling error state if no health detection message is received on the primary port and the secondary port when it is in the complete state, and then switch to the polling error state. After the error state, send an open notification to the secondary port control unit, and periodically send health detection packets on the primary port and the secondary port respectively, if the health detection packets are received on both the primary port and the secondary port, switch to the The above complete state, if only one of the primary port and the secondary port receives a health detection message, then switch to the link single-pass state;

The secondary port control unit is further configured to open the secondary port after receiving the opening notification. the

Preferably, the loop detection unit is further configured to switch to the polling error state if neither the primary port nor the secondary port receives a health detection message when the link is in the one-way state. the

Preferably, the loop detection unit is configured to periodically send a health check on the primary port and the secondary port after detecting that the secondary port has not received a health detection message from the primary port and opening the secondary port. message. the

It can be understood that what is shown in the drawings or the embodiments is only schematic and represents a logical structure, wherein the units shown as separate components may or may not be physically separated, and the components shown as units may be Or it may not be a physical unit, ie it may be located in one place, or distributed over several network elements. the

In addition, the invention also proposes a system for avoiding loops. The system includes a plurality of transmission nodes and master nodes, wherein the transmission nodes can be realized by using the transmission nodes in the prior art, and the master nodes can be realized by any one of the master nodes proposed in the above-mentioned embodiments of the present invention. the

In a word, the above descriptions are only preferred embodiments of the present invention, and are not intended to limit the protection scope of the present invention. All within the spirit and principles of the present invention, any modifications, equivalent replacements, improvements, etc., should be included within the protection scope of the present invention. the

Claims (9)

1. method of avoiding producing loop, it is characterized in that, this method is applied to have the ethernet ring network of loop network structure, comprise host node in this Ethernet ring and more than one transmission node, host node and each transmission node service data transmission on ethernet ring network are finished the miscellaneous service function, and, host node is responsible for the state of each transmission node and link on the ethernet ring network is monitored, and comprising:

Host node on the ethernet ring network sends health detection message on Primary port and secondary port, if only receive health detection message in Primary port and secondary port, and on another, do not receive health detection message, then host node makes secondary port be in blocked state;

Wherein, described host node sends health detection message on Primary port and secondary port step comprises: described host node is in good working condition, secondary port block, host node periodically sends health detection message on master port, if on secondary port, do not receive health detection message, then on secondary port, send health detection message from master port;

Receiving health detection message on described only in Primary port and secondary port is: do not receive the health detection message from master port on secondary port, receive the health detection message from secondary port on master port;

Described host node makes secondary port be in blocked state to comprise: host node keeps the blocked state of secondary port.

2. method according to claim 1, it is characterized in that, only receive after the health detection message in Primary port and secondary port, further comprise: described host node switches to predefined link single-pass state, host node periodically sends health detection message respectively on Primary port and secondary port, if on Primary port and secondary port, all receive health detection message, then switch to described good working condition.

3. method according to claim 2, it is characterized in that, this method further comprises: all do not receive health detection message if be in the described host node of good working condition on Primary port and secondary port, then described host node switches to predefined plooing fault state, the open auxiliary port, and on Primary port and secondary port, periodically send health detection message respectively, if on Primary port and secondary port, all receive health detection message, then switch to described good working condition, if only receive health detection message in Primary port and secondary port, then switch to described link single-pass state, and carry out the processing of blocking secondary port.

4. method according to claim 3, it is characterized in that, when described host node is in link single-pass state, further comprise:, then switch to described plooing fault state if described host node does not all receive health detection message on Primary port and secondary port.

5. host node, it is characterized in that, be arranged in ethernet ring network with loop network structure, comprise host node in this Ethernet ring and more than one transmission node, host node and each transmission node service data transmission on ethernet ring network are finished the miscellaneous service function, and, host node is responsible for the state of each transmission node and link on the ethernet ring network is monitored, and comprising:

The loop detection unit, be used for when switching to good working condition, notify secondary port controlling unit to block secondary port, and periodicity sends health detection message on master port, if on secondary port, do not receive health detection message from master port, then on secondary port, send health detection message, if on master port, receive health detection message from secondary port, only then determine to detect and receive health detection message in Primary port and secondary port, and carry out the processing that sends congestion notification to secondary port controlling unit;

Secondary port controlling unit is used for after receiving congestion notification, makes secondary port be in blocked state.

6. host node according to claim 5, it is characterized in that, described loop detection unit, be further used for when being in good working condition, if on secondary port, do not receive health detection message from master port, on master port, receive health detection message from secondary port, then switch to link single-pass state, after switching to link single-pass state, carry out describedly sending the processing of congestion notification, and further when being in link single-pass state, on Primary port and secondary port, periodically send health detection message respectively to secondary port controlling unit, if on Primary port and secondary port, all receive health detection message, then switch to described good working condition.

7. host node according to claim 6, it is characterized in that, described loop detection unit, be further used for when being in good working condition, if on Primary port and secondary port, all do not receive health detection message, then switch to the plooing fault state, after switching to the plooing fault state, notice is opened in transmission to secondary port controlling unit, and on Primary port and secondary port, periodically send health detection message respectively, if on Primary port and secondary port, all receive health detection message, then switch to described good working condition, if only receive health detection message in Primary port and secondary port, then switch to described link single-pass state;

Described secondary port controlling unit, be further used for receive open notice after, the open auxiliary port.

8. host node according to claim 7 is characterized in that, described loop detection unit is further used for when being in link single-pass state, if all do not receive health detection message on Primary port and secondary port, then switches to described plooing fault state.

9. a system that avoids producing loop is characterized in that, this system comprises a plurality of transmission nodes and as any described host node in the claim 5 to 8.

Images